[db] Faster query for comment garbage collection

Rephrasing a query so it doesn't esceed the statement timeout. Special thanks to @cservakt!
bruntib · Jun 3, 2024 · 1e5d423 · 1e5d423
1 parent e2037fa
commit 1e5d423
Showing 1 changed file with 6 additions and 4 deletions.
diff --git a/web/server/codechecker_server/database/db_cleanup.py b/web/server/codechecker_server/database/db_cleanup.py
@@ -133,12 +133,14 @@ def remove_unused_comments(product):
         LOG.debug("[%s] Garbage collection of dangling comments started...",
                   product.endpoint)
         try:
-            report_hashes = session.query(Report.bug_id) \
-                .group_by(Report.bug_id) \
+            sub = session.query(Comment.id) \
+                .join(Report,
+                      Comment.bug_hash == Report.bug_id,
+                      isouter=True) \
+                .filter(Report.id.is_(None)) \
                 .subquery()
-
             count = session.query(Comment) \
-                .filter(Comment.bug_hash.notin_(report_hashes)) \
+                .filter(Comment.id.in_(sub)) \
                 .delete(synchronize_session=False)
             if count:
                 LOG.debug("%d dangling comments deleted.", count)